Piglet/Finishing Things
The Rant
I've had quite enough of 'vapor ware', projects that will never be finished and even more so those that are doomed from the outset. Thus I thought (with a fair amount of experience of unfinished projects myself) that it was finally time to sort things out. This document probably dosen't come close to solving the problems, but if one more project (even/especially if i'm involved) gets finished as a result of it I'll be very happy.
I'll refer to things a lot in the below bullet points, thus it might be an idea to explain what i mean. Basically a thing means any map, program, feature, mutator or whatever, generally a thing takes some time to do, a fair amount of effort, and a good amount of dedication to complete 'properly'.
Points on finishing things
- Have an idea of what your thing should be. However this is not enough, you must also have a good idea how to make it as well, or it's highly unlikely you'll ever finish it.
- Never attempt a thing you do not currently have the knowledge or materials to see right through to the end. If for somereason you must attempt something and dont have the correct knowledge or materials the get them first!. In many way's this seems obvious, but so many people ignore it that it's worth stating anyway. Theirs nothing more laughable then some idiot with no skill, whose going to make a really good game when its obvious they couldn't make text appear.
- Try to model 'Observentally' instead of 'Properly' (yes they are my own terms before you ask). This is the difference between attempting to model something perfectly, taking everything into account, compared to bodging something that just looks right. It's easier, faster and surprisingly often gives better results. There are many things that this doesn't apply to, however it's still worth a mention.
- Teamwork kills more small projects than it helps (at least in my experience). If you can create your thing by yourself then do so and ignore offers of help. They will change your ideas, meddle with things, argue, annoy and generally worsen the final result. This might seem harsh or unfair and even bad advice. However I have yet to complete a thing with someone else, and believe me, I've tried. Get 'help' on small things but avoid 'partnerships' they reek of faliure.
- Try, Try, Try Again... Give up, come back later, approach it from a different angle, try again then quit. Theres no point getting disheartened.
- Finally and most importantly, always remember A FINISHED BUT POOR THING IS TWICE AS GOOD AS A INCOMPLETE GREAT ONE. And you can quote me on that,
Rant over...
Discussion
Chazums: Interesting little rant, and well said too. A little motto i've come to repeat is "functionality first"
I would note that it is very hard to both stick to one problem and also come back later - i have found this has left me with little pieces missing, but i think this is more down to a lack of a comprihensive requirements spec for what i'm creating than anything else.
Piglet: I'm glad you agree, and that sounds like a good motto. By come back later i originally meant go away and watch some TV etc. but then again working on completly (or as far as possible) different code is often equally good, the important thing is to not get frustrated. Well thanks for the comment 
SabbathCat: Hi Piglet, it all makes a lot of sense, I'd also add "publish your Thing" to the list of crimes. I've seen lots of complete projects never make it to the general public.
It's a common phenomenen, and I admit to sometimes moving onto the latest greatest before shipping out the last thing to be completed. Ever increasing file sizes can hinder the modmaker/skinner/mapper, espcially when it comes to hosting files, but persistence pays off. There are plenty of Unreal/UT/UT2003 mod sites out there for hosting your files, some even let you upload straight on there servers these days with a voting/board system which makes the process at lot easier than it was.
Take the time to package your project with a ReadMe and some background, get it zipped and get it out there.
Foxpaw: I'd have to agree with the above rant.. when I first started building things/programming/what-have-you I had the mindset that I'd do the main stuff now and worry about the details after the "important stuff" was done. My experience has taught me that this does not work. Although you can often find a way to implement the little details afterwards if you are creative, it causes way more headaches.. so I agree wholeheartedly.. leave no detail unfinished, you should know exactly how everything will work before you even start, or you will find yourself redoing a lot of things.
